Allow to modify dinamically training state

[ipmach]

CoreComponent now allow different behaviors of training.

• Default state: it will train until the data gathered is equal to data_use_training.
• Stop sate: it will force to stop the training.
• keep training state: it will force to keep training.

instance = DetectorSomething()

instance.train_state = TrainState.DEFAULT   # (default state)
instance.train_state = TrainState.STOP_TRAINING  # (stop state)
instance.train_state = TrainState.KEEP_TRAINING  # (keep training state)

[viktorbeck98]

Perfect, thanks!